21 May 2014 LWIR polarimetry for enhanced facial recognition in thermal imagery
Author Affiliations +
We present a series of long-wave-infrared (LWIR) polarimetric-based thermal images of facial profiles in which polarization-state information of the image forming radiance is retained and displayed. The resultant polarimetric images show enhanced facial features, additional texture, and details that are not present in the corresponding conventional thermal imagery. It has been generally thought that conventional thermal imagery (MidiR or LWIR) could not produce the detailed spatial information required for reliable human identification due to the so-called "ghosting" effect often seen in thermal imagery of human subjects. By using polarimetric information, we are able to extract subtle surface features of the human face, thus improving subject identification. The considered polarimetric image sets include the conventional thermal intensity image, S0 , the two Stokes images, S1 and S2, and a Stokes image product called the degree-of-linear-polarization (DoLP) image. Finally, Stokes imagery is combined with Fresnel relations to extract additional 3D surface information.
© (2014) COPYRIGHT Society of Photo-Optical Instrumentation Engineers (SPIE). Downloading of the abstract is permitted for personal use only.
Kristan P. Gurton, Kristan P. Gurton, Alex J. Yuffa, Alex J. Yuffa, Gorden Videen, Gorden Videen, } "LWIR polarimetry for enhanced facial recognition in thermal imagery", Proc. SPIE 9099, Polarization: Measurement, Analysis, and Remote Sensing XI, 90990G (21 May 2014); doi: 10.1117/12.2049700; https://doi.org/10.1117/12.2049700

Back to Top